Hadoop is a popular framework for processing large-scale data sets using distributed computing. Many companies use Hadoop to store, manage and analyze their data, and they need skilled professionals who can work with this technology. If you want to prove your Hadoop skills and get certified by LinkedIn, you need to pass the LinkedIn skill assessment test for Hadoop. This test consists of multiple-choice questions that cover various topics related to Hadoop, such as its architecture, components, commands, configuration, and more.

To help you prepare for this test, I have compiled a list of questions and answers that you may encounter in the exam. These questions and answers are based on my own experience and research, and they are not official or endorsed by LinkedIn. However, they can give you an idea of what to expect and how to answer the questions correctly. Here are the questions and answers for the LinkedIn skill assessment test for Hadoop.

Q1. Partitioner controls the partitioning of what data?

•  final keys
•  final values
•  intermediate keys
•  intermediate values

Q2. SQL Windowing functions are implemented in Hive using which keywords?

•  UNION DISTINCT, RANK
•  OVER, RANK
•  OVER, EXCEPT
•  UNION DISTINCT, RANK

Q3. Rather than adding a Secondary Sort to a slow Reduce job, it is Hadoop best practice to perform which optimization?

•  Add a partitioned shuffle to the Map job.
•  Add a partitioned shuffle to the Reduce job.
•  Break the Reduce job into multiple, chained Reduce jobs.
•  Break the Reduce job into multiple, chained Map jobs.

Q4. Hadoop Auth enforces authentication on protected resources. Once authentication has been established, it sets what type of authenticating cookie?

•  encrypted HTTP
•  unsigned HTTP
•  compressed HTTP
•  signed HTTP

Q5. MapReduce jobs can be written in which language?

•  Java or Python
•  SQL only
•  SQL or Java
•  Python or SQL

Q6. To perform local aggregation of the intermediate outputs, MapReduce users can optionally specify which object?

•  Reducer
•  Combiner
•  Mapper
•  Counter

Q7. To verify job status, look for the value ___ in the ___.

•  SUCCEEDED; syslog
•  SUCCEEDED; stdout
•  DONE; syslog
•  DONE; stdout

Q8. Which line of code implements a Reducer method in MapReduce 2.0?

•  public void reduce(Text key, Iterator values, Context context){…}
•  public static void reduce(Text key, IntWritable[] values, Context context){…}
•  public static void reduce(Text key, Iterator values, Context context){…}
•  public void reduce(Text key, IntWritable[] values, Context context){…}

Q9. To get the total number of mapped input records in a map job task, you should review the value of which counter?

•  FileInputFormatCounter
•  FileSystemCounter
•  JobCounter
•  TaskCounter (NOT SURE)

Q10. Hadoop Core supports which CAP capabilities?

•  A, P
•  C, A
•  C, P
•  C, A, P

Q11. What are the primary phases of a Reducer?

•  combine, map, and reduce
•  shuffle, sort, and reduce
•  reduce, sort, and combine
•  map, sort, and combine

Q12. To set up Hadoop workflow with synchronization of data between jobs that process tasks both on disk and in memory, use the ___ service, which is ___.

•  Oozie; open source
•  Oozie; commercial software
•  Zookeeper; commercial software
•  Zookeeper; open source

Q13. For high availability, which type of multiple nodes should you use?

•  data
•  name
•  memory
•  worker

Q14. DataNode supports which type of drives?

•  hot swappable
•  cold swappable
•  warm swappable
•  non-swappable

Q15. Which method is used to implement Spark jobs?

•  on disk of all workers
•  on disk of the master node
•  in memory of the master node
•  in memory of all workers

Q16. In a MapReduce job, where does the map() function run?

•  on the reducer nodes of the cluster
•  on the data nodes of the cluster (NOT SURE)
•  on the master node of the cluster
•  on every node of the cluster

Q17. To reference a master file for lookups during Mapping, what type of cache should be used?

•  distributed cache
•  local cache
•  partitioned cache
•  cluster cache

Q18. Skip bad records provides an option where a certain set of bad input records can be skipped when processing what type of data?

•  cache inputs
•  reducer inputs
•  intermediate values
•  map inputs

Q19. Which command imports data to Hadoop from a MySQL database?

•  spark import –connect jdbc:mysql://mysql.example.com/spark –username spark –warehouse-dir user/hue/oozie/deployments/spark
•  sqoop import –connect jdbc:mysql://mysql.example.com/sqoop –username sqoop –warehouse-dir user/hue/oozie/deployments/sqoop
•  sqoop import –connect jdbc:mysql://mysql.example.com/sqoop –username sqoop –password sqoop –warehouse-dir user/hue/oozie/deployments/sqoop
•  spark import –connect jdbc:mysql://mysql.example.com/spark –username spark –password spark –warehouse-dir user/hue/oozie/deployments/spark

Q20. In what form is Reducer output presented?

•  compressed (NOT SURE)
•  sorted
•  not sorted
•  encrypted

Q21. Which library should be used to unit test MapReduce code?

•  JUnit
•  XUnit
•  MRUnit
•  HadoopUnit

Q22. If you started the NameNode, then which kind of user must you be?

•  hadoop-user
•  super-user
•  node-user
•  admin-user

Q23. State _ between the JVMs in a MapReduce job

•  can be configured to be shared
•  is partially shared
•  is shared
•  is not shared (https://www.lynda.com/Hadoop-tutorials/Understanding-Java-virtual-machines-JVMs/191942/369545-4.html)

Q24. To create a MapReduce job, what should be coded first?

•  a static job() method
•  a Job class and instance (NOT SURE)
•  a job() method
•  a static Job class

Q25. To connect Hadoop to AWS S3, which client should you use?

•  S3A
•  S3N
•  S3
•  the EMR S3

Q26. HBase works with which type of schema enforcement?

•  schema on write
•  no schema
•  external schema
•  schema on read

Q27. HDFS files are of what type?

•  read-write
•  read-only
•  write-only
•  append-only

Q28. A distributed cache file path can originate from what location?

•  hdfs or top
•  http
•  hdfs or http
•  hdfs

Q29. Which library should you use to perform ETL-type MapReduce jobs?

•  Hive
•  Pig
•  Impala
•  Mahout

Q30. What is the output of the Reducer?

•  a relational table
•  an update to the input file
•  a single, combined list
•  a set of <key, value> pairs

map function processes a certain key-value pair and emits a certain number of key-value pairs and the Reduce function processes values grouped by the same key and emits another set of key-value pairs as output.

Q31. To optimize a Mapper, what should you perform first?

•  Override the default Partitioner.
•  Skip bad records.
•  Break up Mappers that do more than one task into multiple Mappers.
•  Combine Mappers that do one task into large Mappers.

Q32. When implemented on a public cloud, with what does Hadoop processing interact?

•  files in object storage
•  graph data in graph databases
•  relational data in managed RDBMS systems
•  JSON data in NoSQL databases

Q33. In the Hadoop system, what administrative mode is used for maintenance?

•  data mode
•  safe mode
•  single-user mode
•  pseudo-distributed mode

Q34. In what format does RecordWriter write an output file?

•  <key, value> pairs
•  keys
•  values
•  <value, key> pairs

Q35. To what does the Mapper map input key/value pairs?

•  an average of keys for values
•  a sum of keys for values
•  a set of intermediate key/value pairs
•  a set of final key/value pairs

Q36. Which Hive query returns the first 1,000 values?

•  SELECT…WHERE value = 1000
•  SELECT … LIMIT 1000
•  SELECT TOP 1000 …
•  SELECT MAX 1000…

Q37. To implement high availability, how many instances of the master node should you configure?

•  one
•  zero
•  shared
•  two or more (https://data-flair.training/blogs/hadoop-high-availability-tutorial)

Q38. Hadoop 2.x and later implement which service as the resource coordinator?

•  kubernetes
•  JobManager
•  JobTracker
•  YARN

Q39. In MapReduce, _ have _

•  tasks; jobs
•  jobs; activities
•  jobs; tasks
•  activities; tasks

Q40. What type of software is Hadoop Common?

•  database
•  distributed computing framework
•  operating system
•  productivity tool

Q41. If no reduction is desired, you should set the numbers of _ tasks to zero.

•  combiner
•  reduce
•  mapper
•  intermediate

Q42. MapReduce applications use which of these classes to report their statistics?

•  mapper
•  reducer
•  combiner
•  counter

Q43. _ is the query language, and _ is storage for NoSQL on Hadoop.

•  HDFS; HQL
•  HQL; HBase
•  HDFS; SQL
•  SQL; HBase

Q44. MapReduce 1.0 _ YARN.

•  does not include
•  is the same thing as
•  includes
•  replaces

Q45. Which type of Hadoop node executes file system namespace operations like opening, closing, and renaming files and directories?

•  ControllerNode
•  DataNode
•  MetadataNode
•  NameNode

Q46. HQL queries produce which job types?

•  Impala
•  MapReduce
•  Spark
•  Pig

Q47. Suppose you are trying to finish a Pig script that converts text in the input string to uppercase. What code is needed on line 2 below?

1 data = LOAD ‘/user/hue/pig/examples/data/midsummer.txt’… 2

•  as (text:CHAR[]); upper_case = FOREACH data GENERATE org.apache.pig.piggybank.evaluation.string.UPPER(TEXT);
•  as (text:CHARARRAY); upper_case = FOREACH data GENERATE org.apache.pig.piggybank.evaluation.string.UPPER(TEXT);
•  as (text:CHAR[]); upper_case = FOREACH data org.apache.pig.piggybank.evaluation.string.UPPER(TEXT);
•  as (text:CHARARRAY); upper_case = FOREACH data org.apache.pig.piggybank.evaluation.string.UPPER(TEXT);

Q48. In a MapReduce job, which phase runs after the Map phase completes?

•  Combiner
•  Reducer
•  Map2
•  Shuffle and Sort

Q49. Where would you configure the size of a block in a Hadoop environment?

•  dfs.block.size in hdfs-site.xmls
•  orc.write.variable.length.blocks in hive-default.xml
•  mapreduce.job.ubertask.maxbytes in mapred-site.xml
•  hdfs.block.size in hdfs-site.xml

Q50. Hadoop systems are _ RDBMS systems.

•  replacements for
•  not used with
•  substitutes for
•  additions for

Q51. Which object can be used to distribute jars or libraries for use in MapReduce tasks?

•  distributed cache
•  library manager
•  lookup store
•  registry

Q52. To view the execution details of an Impala query plan, which function would you use?

•  explain
•  query action
•  detail
•  query plan

Q53. Which feature is used to roll back a corrupted HDFS instance to a previously known good point in time?

•  partitioning
•  snapshot
•  replication
•  high availability

Reference

Q54. Hadoop Common is written in which language?

•  C++
•  C
•  Haskell
•  Java

Q55. Which file system does Hadoop use for storage?

•  NAS
•  FAT
•  HDFS
•  NFS

Q56. What kind of storage and processing does Hadoop support?

•  encrypted
•  verified
•  distributed
•  remote

Q57. Hadoop Common consists of which components?

•  Spark and YARN
•  HDFS and MapReduce
•  HDFS and S3
•  Spark and MapReduce

Q58. Most Apache Hadoop committers’ work is done at which commercial company?

•  Cloudera
•  Microsoft
•  Google
•  Amazon

Q59. To get information about Reducer job runs, which object should be added?

•  Reporter
•  IntReadable
•  IntWritable
•  Writer

Q60. After changing the default block size and restarting the cluster, to which data does the new size apply?

•  all data
•  no data
•  existing data
•  new data

Q61. Which statement should you add to improve the performance of the following query?

SELECT
  c.id,
  c.name,
  c.email_preferences.categories.surveys
FROM customers c;

•  GROUP BY
•  FILTER
•  SUB-SELECT
•  SORT

Q62. What custom object should you implement to reduce IO in MapReduce?

•  Comparator
•  Mapper
•  Combiner
•  Reducer

Q63. You can optimize Hive queries using which method?

•  secondary indices
•  summary statistics
•  column-based statistics
•  a primary key index

Q64. If you are processing a single action on each input, what type of job should you create?

•  partition-only
•  map-only
•  reduce-only
•  combine-only

Q65. The simplest possible MapReduce job optimization is to perform which of these actions?

•  Add more master nodes.
•  Implement optimized InputSplits.
•  Add more DataNodes.
•  Implement a custom Mapper.

Q66. When you implement a custom Writable, you must also define which of these object?

•  a sort policy
•  a combiner policy
•  a compression policy
•  a filter policy

Q67. To copy a file into the Hadoop file system, what command should you use?

•  hadoop fs -copy
•  hadoop fs -copy
•  hadoop fs -copyFromLocal
•  hadoop fs -copyFromLocal

Q68. Delete a Hive _ table and you will delete the table _.

•  managed; metadata
•  external; data and metadata
•  external; metadata
•  managed; data

Q69. To see how Hive executed a JOIN operation, use the _ statement and look for the _ value.

•  EXPLAIN; JOIN Operator
•  QUERY; MAP JOIN Operator
•  EXPLAIN; MAP JOIN Operator
•  QUERY; JOIN Operator

Q70. Pig operates in mainly how many nodes?

•  Two
•  Three
•  Four
•  Five

Q71. After loading data, _ and then run a(n) _ query for interactive queries.

•  invalidate metadata; Impala
•  validate metadata; Impala
•  invalidate metadata; Hive
•  validate metadata; Hive

Q72. In Hadoop MapReduce job code, what must be static?

•  configuration
•  Mapper and Reducer
•  Mapper
•  Reducer

Reference

Q73. In Hadoop simple mode, which object determines the identity of a client process?

•  Kerberos ticket
•  kubernetes token
•  guest operating system
•  host operating system

Reference

Q74. Which is not a valid input format for a MapReduce job?

•  FileReader
•  CompositeInputFormat
•  RecordReader
•  TextInputFormat

Reference

Q75. If you see org.apache.hadoop.mapred, which version of MapReduce are you working with?

•  1.x
•  0.x
•  2.x
•  3.x

Reference